stub resolver timeout
Joseph S D Yao
jsdy at cospo.osis.gov
Wed Jan 5 22:59:09 UTC 2000
----- Forwarded message from ISC Mailing List Manager <listar at isc.org> -----
Date: Mon, 3 Jan 2000 15:56:12 -0500
From: Joseph S D Yao <jsdy at cospo.osis.gov>
To: j j <miffy_bunny at mail.com>
Cc: bind-users at isc.org
Subject: Re: stub resolver timeout
In-Reply-To: <384041241.945102134830.JavaMail.root at web33.pub01>; from miffy_bunny at mail.com on Mon, Dec 13, 1999 at 11:22:14AM -0500
On Mon, Dec 13, 1999 at 11:22:14AM -0500, j j wrote:
> If my DNS client have 2 name server's in its resolv.conf file and the first
> server returned a SERVFAIL to a query, would it try the second name server
> in its resolv.conf for the same query or would it forward the SERVFAIL
> responds back to the application ...
If a resolver can contact ANY server, it will not bother to contact any
other server. Why should it? Each server is supposed to be completely
definitive.
> Another question regarding stub resolvers ...
What are they? I have never heard of them.
> If I have 2 name servers list in my clients resolv.conf file, what is the
> timeout algorithm used by the client. And should I expect the same behavior
> for DNS clients from HP, SUN, Linux, MS and ...
There have been a couple of different algorithms used by the ISC
resolver over the years, which I don't remember. This is what is used
by most Unix and Unix-like systems. Who knows what MS does? Oh, yes:
tools made specially for testing DNS may/will have different behaviour,
e.g., 'nslookup'.
--
Joe Yao jsdy at cospo.osis.gov - Joseph S. D. Yao
COSPO/OSIS Computer Support EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.
----- End forwarded message -----
--
Joe Yao jsdy at cospo.osis.gov - Joseph S. D. Yao
COSPO/OSIS Computer Support EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.
More information about the bind-users
mailing list